Rahdo did a Rahdo Runs Through of 404 today. If you keep hearing me talking about it but don't want to read one of my characteristic billion word posts his run through covers an entire game, referencing most of the rules and his conclusions about the game are insightful and interesting. I'd strongly recommend giving it a look!
The particulars of the demonstration game highlights something unusual though. Rahdo completes a tier four directive (the hardest kind) without lifting a finger. An event frees the monkey which deterministically runs off towards the nearest banana and eats it completing the 'Hide Feeding' objective with no input from any of the players. It begs the question, why is this a difficulty four directive if it can be completed so easily?
This directive is odd but not unique in having a particular combination of traits:
It's easy to complete.
It's easy to interrupt.
There's frequently motivation to interrupt.
An interruption makes it extremely difficult to complete.
There's only one monkey. There are three other missions involving the monkey and all three are easier if the monkey is dead. There are missions involving exposing rooms to space, which kills the monkey. There are missions involving killing scientists, the most action efficient way to kill the scientist in the lab is to expose it to space which also kills the monkey. As a rule of thumb, the monkey dies a lot. Once it's gone, bringing it back requires grabbing the alien artefact (a unique item that is also required by several objectives) or a complex cloning procedure (which is awkward enough that it appears as a directive in its own right.) Those complications aside you also need one of the two bananas on the ship that someone else might have stolen for the other banana related directives, if something gets knocked off course the directive cannot be completed in 6/8ths of the ship (spaces with humans) and if you get it wrong and the monkey eats the banana in the wrong place you have to go and get another one. In general being knocked off course doesn't stop you getting it right next turn with relatively little extra setup time.
All of this adds up to "Sometimes Hide Feeling is really easy, but for the average game it becomes really hard." As mentioned back in the objective difficulty post, the directives are costed for the average game, not their minimum or maximum difficulty. This is an intentional feature of the game, it's supposed to be the case that during the draft players get an idea about the other directives that are in play and can utilise that to spot directives that are going to be particularly easy or hard in the context of a given game. Other features such as the player's starting location and the number of players can also be used to refine these decisions. For instance directives that have a high score because of vulnerabilities to the actions of other robots are inherently better in a game with one other player than they are in a game with five other players. The trick for the player is to recognise the difference between the average difficulty than a directive might present and the difficulty that it's likely to be in a particular game. Perhaps even how hard it will be given the level of aggression typically demonstrated by your friends.
Those are player problems, but this blog is for discussions of game design, so let's talk about the design issue. Ultimately the question is how should the difficulty of the directives be set?
The answer is the same thing that it always is: Playtesting! I started out setting values based on how hard I expected directives to be, usually based on some calculation concerning the number of actions required to complete them under ideal circumstances and some sort of guess about how likely those circumstances were. Some of these estimates were pretty good, others were less good. At the end of the day the only way to tell was to play the game and to see how often different directives were achieved.
I started out with a simple system, sticking a plus next to a directive every time it seemed too easy and a minus each time it seemed too hard - using these to increase or decrease the difficulties of the directives. After a bunch of play, this left directives in one of three camps:
No marks: This directive has the right difficulty. Job done.Several marks of the same type: Right-o, adjust difficulty and test some more.
Several marks of opposite types: Uh-oh
That last category causes all of the headaches and I'm sure it'll be no surprise to you to know that Hide Feeding was there. Another contender in the same category was Destroy Breathing.
The common theme was that directives in this category had situational difficulty. Depending on the number of players, the robots' start locations, the robots' chips and the other directives in play they could vary between very easy and very hard. One solution would have been to remove them, but they were also the directives that were generating the best feedback from playtest groups.
It took me a while to narrow it down, but it came down to a lot of the factors I spelled out above, the fact that they varied in difficulty made the decision to take or pass them more meaningful. Something that might've been a dull, automatic process "Take the card with the lowest number." became an interesting strategic decision "I know destroy war went by so if I expose science and cloning to space, the player with that will probably expose weaponry and navigation for me." Players loved it and I didn't want to take that away from them.
It did leave me with a pile of objectives that had no clear difficulty number associated with them though. I could reasonably rate any of them one through four inclusive, as that number would be correct in some proportion of games played. Ultimately I let three considerations guide the decision:
1) Playtesting was still important, if directives were completed more or less often that's still a guide to overall difficulty - even if it might not be appropriate to a specific individual game.
2) The difficulty numbers partially serve to help newbies make decisions. If it was close I kicked them to the higher difficulty bracket to help new players avoid cards that required more complicated judgements.
3) Fun is king. Ultimately players enjoyed games where someone picked a high difficulty objective and found a way to complete it easily much more than ones where someone picked a low difficulty objective and found it unexpectedly hard. So if things were close it was another reason to kick them up a difficulty rating.
A thousand words in and I'm ready to get to the point: I think it's part of a wider pattern of games doing well because they contain things with a nominal value that does not necessarily reflect their real value. More on this later in the week!
For now, go and back 404: Law Not Found, if you haven't already. It's good, Rahdo liked it and everything